Hierarchical Semi-Implicit Variational Inference with Application to Diffusion Model Acceleration
Peking University（北京大学）；Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所）；Beijing Academy of Artificial Intelligence（北京人工智能研究院）；Megvii Technology Inc.（旷视科技有限公司）
Semi-implicit variational inference (SIVI) has been introduced to expand the analytical variational families by defining expressive semi-implicit distributions in a hierarchical manner. However, the single-layer architecture commonly used in current SIVI methods can be insufficient when the target posterior has complicated structures. In this paper, we propose hierarchical semi-implicit variational inference, called HSIVI, which generalizes SIVI to allow more expressive multi-layer construction of semi-implicit distributions. By introducing auxiliary distributions that interpolate between a simple base distribution and the target distribution, the conditional layers can be trained by progressively matching these auxiliary distributions one layer after another. Moreover, given pre-trained score networks, HSIVI can be used to accelerate the sampling process of diffusion models with the score matching objective. We show that HSIVI significantly enhances the expressiveness of SIVI on several Bayesian inference problems with complicated target distributions. When used for diffusion model acceleration, we show that HSIVI can produce high quality samples comparable to or better than the existing fast diffusion model based samplers with a small number of function evaluations on various datasets.
Columbia University（哥伦比亚大学）；Princeton University（普林斯顿大学）
Implicit probabilistic models are a flexible class of models defined by a simulation process for data. They form the basis for theories which encompass our understanding of the physical world. Despite this fundamental nature, the use of implicit models remains limited due to challenges in specifying complex latent structure in them, and in performing inferences in such models with large data sets. In this paper, we first introduce hierarchical implicit models (HIMs). HIMs combine the idea of implicit densities with hierarchical Bayesian modeling, thereby defining models via simulators of data …
Texas A&M University（得克萨斯农工大学）；The University of Texas at Austin（得克萨斯大学奥斯汀分校）
Semi-implicit graph variational auto-encoder (SIG-VAE) is proposed to expand the flexibility of variational graph auto-encoders (VGAE) to model graph data. SIG-VAE employs a hierarchical variational framework to enable neighboring node sharing for better generative modeling of graph dependency structure, together with a Bernoulli-Poisson link decoder. Not only does this hierarchical construction provide a more flexible generative graph model to better capture real-world graph properties, but also does SIG-VAE naturally lead to semi-implicit hierarchical variational inference that allows faithf…
Efficient modeling of relational data arising in physical, social, and information sciences is challenging due to complicated dependencies within the data. In this work, we build off of semi-implicit graph variational auto-encoders to capture higher-order statistics in a low-dimensional graph latent representation. We incorporate hyperbolic geometry in the latent space through a Poincare embedding to efficiently represent graphs exhibiting hierarchical structure. To address the naive posterior latent distribution assumptions in classical variational inference, we use semi-implicit hierarchical…
To combine explicit and implicit generative models, we introduce semi-implicit generator (SIG) as a flexible hierarchical model that can be trained in the maximum likelihood framework. Both theoretically and experimentally, we demonstrate that SIG can generate high quality samples especially when dealing with multi-modality. By introducing SIG as an unbiased regularizer to the generative adversarial network (GAN), we show the interplay between maximum likelihood and adversarial learning can stabilize the adversarial training, resist the notorious mode collapsing problem of GANs, and improve th…
Athens University of Economics and Business（雅典经济与商业大学）；University of Cambridge（剑桥大学）；Columbia University（哥伦比亚大学）
We develop unbiased implicit variational inference (UIVI), a method that expands the applicability of variational inference by defining an expressive variational family. UIVI considers an implicit variational distribution obtained in a hierarchical manner using a simple reparameterizable distribution whose variational parameters are defined by arbitrarily flexible deep neural networks. Unlike previous works, UIVI directly optimizes the evidence lower bound (ELBO) rather than an approximation to the ELBO. We demonstrate UIVI on several models, including Bayesian multinomial logistic regression …
How to obtain hierarchical representations with an increasing level of abstraction becomes one of the key issues of learning with deep neural networks. A variety of RNN models have recently been proposed to incorporate both explicit and implicit hierarchical information in modeling languages in the literature. In this paper, we propose a novel approach called the latent indicator layer to identify and learn implicit hierarchical information (e.g., phrases), and further develop an EM algorithm to handle the latent indicator layer in training. The latent indicator layer further simplifies a text…
Mila, University of Montreal（蒙特利尔大学Mila）；Element AI；CIFAR（加拿大高等研究院）
Importance weighted variational inference (Burda et al., 2015) uses multiple i.i.d. samples to have a tighter variational lower bound. We believe a joint proposal has the potential of reducing the number of redundant samples, and introduce a hierarchical structure to induce correlation. The hope is that the proposals would coordinate to make up for the error made by one another to reduce the variance of the importance estimator. Theoretically, we analyze the condition under which convergence of the estimator variance can be connected to convergence of the lower bound. Empirically, we confirm t…
We analyze the learning dynamics of infinitely wide neural networks with a finite sized bottle-neck. Unlike the neural tangent kernel limit, a bottleneck in an otherwise infinite width network al-lows data dependent feature learning in its bottle-neck representation. We empirically show that a single bottleneck in infinite networks dramatically accelerates training when compared to purely in-finite networks, with an improved overall performance. We discuss the acceleration phenomena by drawing similarities to infinitely wide deep linear models, where the acceleration effect of a bottleneck can…
Princeton University（普林斯顿大学）；Institute for Advanced Study（高等研究院）；Google Brain（谷歌大脑）
Conventional wisdom in deep learning states that increasing depth improves expressiveness but complicates optimization. This paper suggests that, sometimes, increasing depth can speed up optimization. The effect of depth on optimization is decoupled from expressiveness by focusing on settings where additional layers amount to overparameterization - linear neural networks, a well-studied model. Theoretical analysis, as well as experiments, show that here depth acts as a preconditioner which may accelerate convergence. Even on simple convex problems such as linear regression with $\ell_p$ loss, …
In the pursuit of explaining implicit regularization in deep learning, prominent focus was given to matrix and tensor factorizations, which correspond to simplified neural networks. It was shown that these models exhibit an implicit tendency towards low matrix and tensor ranks, respectively. Drawing closer to practical deep learning, the current paper theoretically analyzes the implicit regularization in hierarchical tensor factorization, a model equivalent to certain deep convolutional neural networks. Through a dynamical systems lens, we overcome challenges associated with hierarchy, and est…
In this paper, we propose Hierarchical Federated Learning with Momentum Acceleration (HierMo), a three-tier worker-edge-cloud federated learning algorithm that applies momentum for training acceleration. Momentum is calculated and aggregated in the three tiers. We provide convergence analysis for HierMo, showing a convergence rate of O(1/T). In the analysis, we develop a new approach to characterize model aggregation, momentum aggregation, and their interactions. Based on this result, {we prove that HierMo achieves a tighter convergence upper bound compared with HierFAVG without momentum}. We …
University of Bonn（波恩大学）；University of Oxford（牛津大学）；Lamarr Institute for Machine Learning and Artificial Intelligence（拉玛尔机器学习与人工智能研究所）
Accurate mapping of large-scale environments is an essential building block of most outdoor autonomous systems. Challenges of traditional mapping methods include the balance between memory consumption and mapping accuracy. This paper addresses the problem of achieving large-scale 3D reconstruction using implicit representations built from 3D LiDAR measurements. We learn and store implicit features through an octree-based, hierarchical structure, which is sparse and extensible. The implicit features can be turned into signed distance values through a shallow neural network. We leverage binary c…
NVIDIA（英伟达）；University of Waterloo（滑铁卢大学）；University of Toronto（多伦多大学）；Vector Institute（矢量研究所）
We present HIPNet, a neural implicit pose network trained on multiple subjects across many poses. HIPNet can disentangle subject-specific details from pose-specific details, effectively enabling us to retarget motion from one subject to another or to animate between keyframes through latent space interpolation. To this end, we employ a hierarchical skeleton-based representation to learn a signed distance function on a canonical unposed space. This joint-based decomposition enables us to represent subtle details that are local to the space around the body joint. Unlike previous neural implicit …
Department of Computer Science, The University of Chicago（计算机科学系，芝加哥大学）；Department of Geophysical Sciences, The University of Chicago（地球科学系，芝加哥大学）；Department of Mechanical Engineering, Rice University（机械工程系， Rice大学）；Department of Physics, The University of Chicago（物理学系，芝加哥大学）；Department of Electrical and Computer Engineering, Tufts University（电气与计算机工程系， Tufts大学）；Department of Statistics, The University of Chicago（统计学系，芝加哥大学）
Università degli Studi di Genova（热那亚大学）；Machine Learning Genoa Center（热那亚机器学习中心）；Massachusetts Institute of Technology（麻省理工学院）；Istituto Italiano di Tecnologia（意大利技术研究院）
We study learning properties of accelerated gradient descent methods for linear least-squares in Hilbert spaces. We analyze the implicit regularization properties of Nesterov acceleration and a variant of heavy-ball in terms of corresponding learning error bounds. Our results show that acceleration can provides faster bias decay than gradient descent, but also suffers of a more unstable behavior. As a result acceleration cannot be in general expected to improve learning accuracy with respect to gradient descent, but rather to achieve the same accuracy with reduced computations. Our theoretical…
Massachusetts Institute of Technology (MIT)（麻省理工学院）；The University of Maryland（马里兰大学）；Harvard University（哈佛大学）
Graph Neural Networks (GNNs) have been studied through the lens of expressive power and generalization. However, their optimization properties are less well understood. We take the first step towards analyzing GNN training by studying the gradient dynamics of GNNs. First, we analyze linearized GNNs and prove that despite the non-convexity of training, convergence to a global minimum at a linear rate is guaranteed under mild assumptions that we validate on real-world graphs. Second, we study what may affect the GNNs' training speed. Our results show that the training of GNNs is implicitly accel…
ETH Zürich（苏黎世联邦理工学院）；INRIA Paris（法国国家信息与自动化研究所巴黎分部）
Viewing optimization methods as numerical integrators for ordinary differential equations (ODEs) provides a thought-provoking modern framework for studying accelerated first-order optimizers. In this literature, acceleration is often supposed to be linked to the quality of the integrator (accuracy, energy preservation, symplecticity). In this work, we propose a novel ordinary differential equation that questions this connection: both the explicit and the semi-implicit (a.k.a symplectic) Euler discretizations on this ODE lead to an accelerated algorithm for convex programming. Although semi-imp…
University of California, Berkeley（加州大学伯克利分校）；Institute for Advanced Study（普林斯顿高等研究院）；University of Pennsylvania（宾夕法尼亚大学）
We study first-order optimization methods obtained by discretizing ordinary differential equations (ODEs) corresponding to Nesterov's accelerated gradient methods (NAGs) and Polyak's heavy-ball method. We consider three discretization schemes: an explicit Euler scheme, an implicit Euler scheme, and a symplectic scheme. We show that the optimization algorithm generated by applying the symplectic scheme to a high-resolution ODE proposed by Shi et al. [2018] achieves an accelerated rate for minimizing smooth strongly convex functions. On the other hand, the resulting algorithm either fails to ach…
Samsung AI Center Moscow（三星莫斯科人工智能中心）；National Research University Higher School of Economics（国立研究型高等经济学院）；NRU HSE（国立研究型高等经济学院）
Variational Inference is a powerful tool in the Bayesian modeling toolkit, however, its effectiveness is determined by the expressivity of the utilized variational distributions in terms of their ability to match the true posterior distribution. In turn, the expressivity of the variational family is largely limited by the requirement of having a tractable density function. To overcome this roadblock, we introduce a new family of variational upper bounds on a marginal log density in the case of hierarchical models (also known as latent variable models). We then give an upper bound on the Kullba…
Hierarchical Multi-Relational Graph Representation Learning for Large-Scale Prediction of Drug-Drug Interactions
School of Electronics and Information Engineering, Xi’an Jiaotong University（西安交通大学电子与信息工程学院）；College of Geomatics, Xi’an University of Science and Technology（西安科技大学测绘学院）；Aerospace Information Research Institute, Chinese Academy of Sciences（中国科学院空天信息创新研究院）
University of Science & Technology of China（中国科学技术大学）；Microsoft Research Asia（微软亚洲研究院）；University of Illinois at Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）
The momentum acceleration technique is widely adopted in many optimization algorithms. However, there is no theoretical answer on how the momentum affects the generalization performance of the optimization algorithms. This paper studies this problem by analyzing the implicit regularization of momentum-based optimization. We prove that on the linear classification problem with separable data and exponential-tailed loss, gradient descent with momentum (GDM) converges to the L2 max-margin solution, which is the same as vanilla gradient descent. That means gradient descent with momentum accelerati…
Collaborative Metric Learning (CML) recently emerged as a powerful paradigm for recommendation based on implicit feedback collaborative filtering. However, standard CML methods learn fixed user and item representations, which fails to capture the complex interests of users. Existing extensions of CML also either ignore the heterogeneity of user-item relations, i.e. that a user can simultaneously like very different items, or the latent item-item relations, i.e. that a user's preference for an item depends, not only on its intrinsic characteristics, but also on items they previously interacted …
University of Pennsylvania（宾夕法尼亚大学）；Stanford University（斯坦福大学）
Acceleration and momentum are the de facto standard in modern applications of machine learning and optimization, yet the bulk of the work on implicit regularization focuses instead on unaccelerated methods. In this paper, we study the statistical risk of the iterates generated by Nesterov's accelerated gradient method and Polyak's heavy ball method, when applied to least squares regression, drawing several connections to explicit penalization. We carry out our analyses in continuous-time, allowing us to make sharper statements than in prior work, and revealing complex interactions between earl…
East China Normal University（华东师范大学）；Microsoft Research Asia（微软亚洲研究院）
Generative graph self-supervised learning (SSL) aims to learn node representations by reconstructing the input graph data. However, most existing methods focus on unsupervised learning tasks only and very few work has shown its superiority over the state-of-the-art graph contrastive learning (GCL) models, especially on the classification task. While a very recent model has been proposed to bridge the gap, its performance on unsupervised learning tasks is still unknown. In this paper, to comprehensively enhance the performance of generative graph SSL against other GCL models on both unsupervise…
University of California, Berkeley（加州大学伯克利分校）；University of California, Davis（加州大学戴维斯分校）
We apply the technique of implicit differentiation to boost performance, reduce numerical error, and remove required user-tuning in the Marginal Unbiased Score Expansion (MUSE) algorithm for hierarchical Bayesian inference. We demonstrate these improvements on three representative inference problems: 1) an extended Neal's funnel 2) Bayesian neural networks, and 3) probabilistic principal component analysis. On our particular test cases, MUSE with implicit differentiation is faster than Hamiltonian Monte Carlo by factors of 155, 397, and 5, respectively, or factors of 65, 278, and 1 without imp…
